The success of autologous marrow transplantation (AMT) for the treatment of patients with leukemia or lymphoma is limited largely by the high incidence of recurrence of the malignancy after AMT. Immunotherapy in the form of recombinant human IL-2 used alone or with ex vivo-generated lymphokine activated killer (LAK) cells has induced regression of advanced cancer in some patients and remissions in some patients with acute leukemia or malignant lymphoma. IL-2/LAK therapy could potentially represent a treatment modality which would not be cross-resistant with the chemoradiotherapy used as a preparative regimen for AMT. Our goal is to test the hypothesis that this approach, used early after AMT in a setting of minimal residual disease, i.e., as consolidative immunotherapy, may decrease the relapse rates. We have already (a) demonstrated that IL-2-responsive LAK precursor cells are detectable in the circulation early after AMT; (b) identified an IL-2 regimen which is tolerable and induces immunomodulatory effects when administered early after AMT; and (c) completed a feasibility trial of the IL-2 regimen plus ex vivo-generated autologous LAK cells in patients who have undergone AMT for advanced hematologic malignancy -- with encouraging preliminary clinical results. The specific aim of this proposal is to perform a randomized controlled trial of consolidative immunotherapy with IL-2 + LAK cells vs. no IL-2/LAK after AMT in patients with non-Hodgkin lymphoma and acute lymphoblastic leukemia so as to determine the effect, if any, on their high relapse rates.